Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. An estimation results display system comprising: an input unit, implemented by an input device, for receiving input of an estimation result data set including information associated with an estimation result and information indicating a learning model used when deriving the estimation result; an estimation unit, implemented by the processor, for selecting the learning model based on at least one attribute value included in the input estimation result data set; and a display unit, implemented by a processor and a display device, for displaying a graph that represents the estimation result by a symbol and, in the case where any symbol in the graph is selected, displaying information indicating a learning model corresponding to an estimation result represented by the selected symbol; wherein the estimation result received by the input unit is derived using estimation data and the learning model selected depending on a selection model being a tree structure model in which each leaf node is a learning model and each node other than the leaf nodes is a condition relating to the estimation data.
2. The estimation results display system according to claim 1 , wherein the estimation result received by the input unit is derived using estimation data and the learning model selected depending on the estimation data.
The system relates to a display system for estimation results, particularly in applications where different types of estimation data require different learning models for accurate analysis. The problem addressed is the need for a flexible system that can adapt to various types of input data by selecting the appropriate learning model, ensuring accurate and reliable estimation results. The system includes an input unit that receives estimation results derived from estimation data and a selected learning model. The learning model is chosen based on the type or characteristics of the estimation data to optimize the estimation process. For example, if the estimation data involves time-series analysis, a recurrent neural network may be selected, while for image-based data, a convolutional neural network might be preferred. The system ensures that the estimation results are displayed in a user-friendly manner, allowing for easy interpretation and decision-making. The display system may also include a user interface that allows users to input or modify estimation data, select or adjust learning models, and view the estimation results. The system dynamically adapts to the input data, ensuring that the most suitable learning model is applied, thereby improving the accuracy and reliability of the estimation results. This adaptability makes the system versatile for various applications, including predictive analytics, quality control, and decision support systems.
3. The estimation results display system according to claim 1 , further comprising an estimation unit, implemented by the processor, for selecting the learning model corresponding to a condition satisfied by estimation data from among a plurality of learning models, deriving the estimation result based on the estimation data and the selected learning model, and outputting the information associating the estimation result and the information indicating the learning model used when deriving the estimation result, wherein the input unit receives the information output by the estimation unit.
This invention relates to a system for displaying estimation results derived from machine learning models. The system addresses the challenge of selecting an appropriate learning model for a given set of estimation data and providing transparent, traceable results. The system includes an estimation unit that selects a learning model from a plurality of models based on conditions satisfied by the input estimation data. The selected model is used to derive an estimation result, and the system outputs both the result and metadata indicating which model was used. This metadata includes information about the model's parameters, structure, or training data, ensuring traceability and reproducibility. The system also includes an input unit that receives the output from the estimation unit, allowing users to review the model selection process and the derived results. The display system ensures that users can understand the basis for each estimation, improving reliability and trust in the outputs. The invention is particularly useful in applications where model selection and result transparency are critical, such as financial forecasting, medical diagnostics, or industrial quality control.
4. The estimation results display system according to claim 1 , wherein the display unit displays the graph that represents a value of the estimation result by a coordinate at which the symbol is located.
This system relates to a display system for presenting estimation results, particularly in fields such as data analysis, predictive modeling, or monitoring systems where visual representation of estimated values is critical. The problem addressed is the need for an intuitive and accurate way to display estimation results, ensuring users can quickly interpret the data without ambiguity. The system includes a display unit that presents a graph representing the value of an estimation result. The graph uses a coordinate system where a symbol is placed to indicate the estimated value. The symbol's position on the graph corresponds directly to the numerical or categorical value of the estimation result, allowing users to visually assess the result by observing the symbol's location. This approach enhances clarity and reduces the need for additional annotations or legends, making the data more accessible. The display unit may also include features such as dynamic updates, allowing the graph to reflect real-time or near-real-time changes in estimation results. The system may further incorporate user interaction, enabling users to adjust the graph's scale, view historical data, or compare multiple estimation results simultaneously. The use of symbols in a coordinate-based graph ensures that the estimation results are presented in a structured and easily interpretable manner, improving decision-making processes in applications such as financial forecasting, industrial monitoring, or scientific research.
5. The estimation results display system according to claim 1 , wherein the input unit receives input of a set of estimation result data which is information associating a value of the estimation result, the information indicating the learning model used when deriving the estimation result, and an actual measured value corresponding to the estimation result, and wherein the display unit displays a scatter graph that has an axis corresponding to the value of the estimation result and an axis corresponding to the actual measured value and in which the symbol corresponding to the estimation result data is placed based on the value of the estimation result and the actual measured value corresponding to the estimation result data.
This invention relates to a system for visualizing estimation results from machine learning models. The system addresses the challenge of evaluating model performance by providing a clear graphical representation of how well estimated values align with actual measured values. The system includes an input unit that receives a dataset containing estimation results, the learning model used to generate those results, and corresponding actual measured values. A display unit generates a scatter plot where one axis represents the estimated values and the other axis represents the actual measured values. Each data point on the plot corresponds to a pair of estimated and actual values, allowing users to visually assess the accuracy and reliability of the model's predictions. The system helps users quickly identify patterns, biases, or discrepancies between estimated and actual values, facilitating model evaluation and improvement. The scatter plot may include symbols or markers to distinguish different models or datasets, enhancing interpretability. This visualization tool is particularly useful in fields where model performance is critical, such as predictive analytics, quality control, and scientific research.
6. The estimation results display system according to claim 1 , wherein the display unit displays the graph that represents the estimation result by a type of the symbol.
This system relates to a display system for presenting estimation results, particularly in fields such as data analysis, predictive modeling, or monitoring systems where visual representation of estimated outcomes is critical. The problem addressed is the need for clear, intuitive visualization of estimation results to improve user comprehension and decision-making. Traditional systems often rely on generic graphs or tables, which may not effectively convey the nuances of different estimation types or their reliability. The system includes a display unit that generates a graph representing estimation results, where each type of estimation result is visually distinguished by a unique symbol. This allows users to quickly identify and interpret different estimation categories or confidence levels at a glance. The symbols may vary in shape, color, or other visual attributes to enhance clarity. Additionally, the system may incorporate a method for dynamically adjusting the graph based on user input or real-time data updates, ensuring the display remains relevant and accurate. The use of distinct symbols helps reduce cognitive load and improves the efficiency of data interpretation, particularly in complex or time-sensitive scenarios. This approach is particularly useful in applications such as financial forecasting, medical diagnostics, or industrial process monitoring, where rapid and accurate understanding of estimation results is essential.
7. The estimation results display system according to claim 6 , wherein the input unit receives input of a set of estimation result data which is information associating the estimation result, values of at least two types of attributes used when deriving the estimation result, and the information indicating the learning model used when deriving the estimation result, and wherein the display unit displays a scatter graph that has an axis corresponding to a value of a first attribute of the attributes and an axis corresponding to a value of a second attribute of the attributes and in which the symbol corresponding to the estimation result data is placed based on the value of the first attribute and the value of the second attribute corresponding to the estimation result data.
This invention relates to a system for displaying estimation results derived from machine learning models. The system addresses the challenge of visualizing and analyzing estimation results in a way that allows users to understand the relationships between different input attributes and the model's predictions. The system includes an input unit that receives a set of estimation result data, which includes the estimation result itself, values of at least two types of attributes used to derive the result, and information about the learning model used. The display unit generates a scatter graph where each axis represents a different attribute value, and symbols corresponding to the estimation result data are plotted based on their attribute values. This visualization helps users identify patterns, correlations, or anomalies in the estimation results relative to the input attributes. The system may also allow users to filter or adjust the displayed data to focus on specific subsets of results or models. By providing a clear graphical representation of how different attributes influence estimation outcomes, the system enhances interpretability and decision-making in machine learning applications.
8. The estimation results display system according to claim 1 , wherein the display unit groups learning models and displays a line graph that represents a change of the estimation result and in which an attribute of a line is changed depending on each learning model group.
This invention relates to a system for displaying estimation results from multiple learning models, addressing the challenge of visualizing and comparing performance across different models in a clear and organized manner. The system groups learning models into categories and presents their estimation results as a line graph, where each group is visually distinguished by varying line attributes such as color, style, or thickness. This allows users to easily track changes in estimation accuracy or other metrics over time and compare trends between different model groups. The display unit dynamically adjusts the line attributes to ensure clear differentiation, enhancing interpretability. The system may also include features for selecting specific models or groups for detailed analysis, filtering results based on criteria, and adjusting the graph's time range or scale. By organizing models into groups and applying distinct visual attributes, the system simplifies the evaluation of model performance, making it easier to identify patterns, outliers, or improvements in estimation accuracy. The invention is particularly useful in fields like machine learning, data analytics, and predictive modeling, where comparing multiple models is essential for optimization and decision-making.
9. An estimation results display method comprising: receiving input of an estimation result data set including information associated with an estimation result and information indicating a learning model used when deriving the estimation result; selecting the learning model based on at least one attribute value included in the input estimation result data set; and displaying a graph that represents the estimation result by a symbol and, in the case where any symbol in the graph is selected, displaying information indicating a learning model corresponding to an estimation result represented by the selected symbol; wherein the estimation result that is received is derived using estimation data and the learning model selected depending on a selection model being a tree structure model in which each leaf node is a learning model and each node other than the leaf nodes is a condition relating to the estimation data.
This invention relates to a method for displaying estimation results derived from machine learning models, addressing the challenge of effectively presenting and interpreting complex model outputs. The method involves receiving an estimation result data set that includes both the estimation result and metadata indicating the specific learning model used to generate it. The system then selects the appropriate learning model based on attribute values within the input data set. A graph is displayed, where each estimation result is represented by a symbol. When a user selects a symbol, the system displays detailed information about the corresponding learning model, including its structure and parameters. The estimation results are derived using a hierarchical tree structure model, where each leaf node represents a distinct learning model, and non-leaf nodes contain conditions that determine which model to apply based on the input estimation data. This approach allows users to trace the origin of each estimation result back to the specific model and conditions that produced it, enhancing transparency and interpretability in machine learning workflows. The method is particularly useful in applications where multiple models are deployed dynamically, and users need to understand the reasoning behind each prediction or estimation.
10. A non-transitory computer-readable recording medium in which an estimation results display program is recorded, the estimation results display program provided in a computer including an input unit for receiving input of an estimation result data set including information associated with an estimation result and information indicating a learning model used when deriving the estimation result, the computer including an estimation unit for selecting the learning model based on at least one attribute value included in the input estimation result data set, the estimation results display program causing the computer to execute a display process of displaying a graph that represents the estimation result by a symbol and, in the case where any symbol in the graph is selected, displaying information indicating a learning model corresponding to an estimation result represented by the selected symbol; wherein the estimation result received by the input unit is derived using estimation data and the learning model selected depending on a selection model being a tree structure model in which each leaf node is a learning model and each node other than the leaf nodes is a condition relating to the estimation data.
This invention relates to a system for displaying estimation results derived from machine learning models, addressing the challenge of effectively presenting and interpreting results from multiple models in a structured manner. The system involves a computer-readable medium storing a program that processes estimation result data sets, which include both the estimation results and metadata indicating the specific learning model used to generate them. The program executes a display process that visualizes the results as a graph, where each data point is represented by a symbol. When a user selects a symbol, the system reveals the corresponding learning model that produced that result. The estimation results are derived using a hierarchical selection model structured as a decision tree, where each leaf node represents a distinct learning model and non-leaf nodes contain conditions that determine which model is applied based on input data attributes. This approach allows users to trace the origin of each result back to its underlying model, enhancing transparency and interpretability in machine learning workflows. The system is particularly useful in scenarios where multiple models are deployed to handle different subsets of data, ensuring clarity in how results are generated.
11. An estimation results display system comprising: an input unit, implemented by an input device, for receiving input of an estimation result data set including information associated with an estimation result and information indicating a learning model used when deriving the estimation result; an estimation unit, implemented by the processor, for selecting the learning model based on at least one attribute value included in the input estimation result data set; and a display unit, implemented by a processor and a display device, for displaying a graph that represents the estimation result by a symbol and, in the case where any of a plurality of learning models is selected, displaying a symbol representing an estimation result derived using the selected learning model in a mode different from other symbols; wherein the estimation result received by the input unit is derived using estimation data and the learning model selected depending on a selection model being a tree structure model in which each leaf node is a learning model and each node other than the leaf nodes is a condition relating to the estimation data.
This system relates to visualizing estimation results derived from machine learning models, addressing the challenge of interpreting and comparing outputs from multiple models. The system receives an estimation result data set containing the result itself and metadata about the learning model used to generate it. A processor selects the appropriate learning model based on attribute values in the input data, using a tree-structured selection model where each leaf node represents a learning model and non-leaf nodes contain conditions for selecting those models. The system then displays the estimation results as symbols on a graph, differentiating symbols for results derived from different models by using distinct visual modes (e.g., color, shape, or size). This allows users to easily identify which model produced each result, aiding in model performance analysis and comparison. The system is particularly useful in scenarios where multiple models are applied to the same estimation data, requiring clear visualization of their outputs. The tree-structured selection model ensures that the correct model is highlighted in the display, providing context for the estimation results.
12. An estimation results display method comprising: selecting a learning model based on at least one attribute value included in an input estimation data set; receiving input of the estimation result data set including information associated with an estimation result and information indicating the learning model used when deriving the estimation result; and displaying a graph that represents the estimation result by a symbol and, in the case where any of a plurality of learning models is selected, displaying a symbol representing an estimation result derived using the selected learning model in a mode different from other symbols; wherein the estimation result that is received is derived using estimation data and the learning model selected depending on a selection model being a tree structure model in which each leaf node is a learning model and each node other than the leaf nodes is a condition relating to the estimation data.
This technical summary describes a method for displaying estimation results derived from multiple learning models. The method addresses the challenge of visualizing and comparing estimation results when different learning models are used, ensuring clarity in distinguishing between results obtained from different models. The system selects a learning model based on attribute values in an input estimation dataset, then receives estimation result data, which includes the result itself and metadata indicating the learning model used. The method displays a graph where each estimation result is represented by a symbol, with results from different models shown in distinct modes (e.g., color, shape, or size) to avoid confusion. The estimation results are derived using a selection model structured as a tree, where each leaf node corresponds to a learning model and non-leaf nodes represent conditions applied to the estimation data. This hierarchical approach ensures that the appropriate model is chosen based on the input data's attributes, improving the accuracy and interpretability of the results. The method enhances transparency in model-based estimations by clearly differentiating results from different models in the visualization.
13. A non-transitory computer-readable recording medium in which an estimation results display program is recorded, the estimation results display program provided in a computer including an input unit for receiving input of an estimation result data set including information associated with an estimation result and information indicating a learning model used when deriving the estimation result, the computer including an estimation unit for selecting the learning model based on at least one attribute value included in the input estimation result data set, the estimation results display program causing the computer to execute a display process of displaying a graph that represents the estimation result by a symbol and, in the case where any of a plurality of learning models is selected, displaying a symbol representing an estimation result derived using the selected learning model in a mode different from other symbols; wherein the estimation result received by the input unit is derived using estimation data and the learning model selected depending on a selection model being a tree structure model in which each leaf node is a learning model and each node other than the leaf nodes is a condition relating to the estimation data.
This invention relates to a system for visualizing estimation results derived from machine learning models, addressing the challenge of effectively presenting and distinguishing results obtained from different models. The system includes a computer-readable medium storing a program that processes and displays estimation results, where each result is associated with metadata indicating the learning model used for its derivation. The program causes the computer to display a graph where each estimation result is represented by a symbol, with results from different models visually differentiated. When multiple models are used, the system highlights symbols corresponding to results from a selected model in a distinct mode (e.g., color, shape, or size) compared to others. The estimation results are generated using a hierarchical selection model—a tree structure where leaf nodes represent individual learning models and non-leaf nodes represent conditions applied to the input data to determine which model to use. This approach ensures clarity in tracking which model produced each result, aiding in model evaluation and debugging. The system is particularly useful in scenarios where multiple models are deployed, such as in adaptive or ensemble learning systems.
Unknown
January 5, 2021
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.